Clinical endpoint sensitivity in rheumatoid arthritis: modeling and simulation.

The commonly used efficacy endpoints in Rheumatoid Arthritis (RA) clinical trials are American College of Rheumatology 20 % improvement criteria (ACR20), ACR50, and ACR70 response rates, and the 28-joint disease activity score (DAS28). Longitudinal models to quantitate the exposure-response relationships for ACRs and DAS28 score were developed for four biologics used for the management of RA. The models were then used to simulate the clinical outcome at various time points following different treatment regimens. Discriminative sensitivity of these endpoints was assessed using a power analysis. The trial simulation and subsequent power analysis showed that both ACR20 and DAS28 exhibit much lower power in distinguishing between two doses investigated compared with distinguishing treatment effect over placebo/Methotrexate (MTX) control. ACR20 response rate is generally more powerful in detecting treatment effect over placebo/MTX control as compared to DAS28. The findings of current study provide useful information which will help future clinical trial design for the treatment of patients with RA.